OK, OK, so I know all the arguments. I've never participated in a DBT, but I might like to do something like that in my own home. I would like to compare two systems, and I would like to be able to rather instantly switch between the two. Suppose two systems:
DMP-BD30-> old Denon 170x receiver->Vandersteen
DMP-BD30->Cary Cinema 11->Sunfine Cinema Grand->Vandersteen
So we'll basically be comparing DACs, preamps and amps all in one shot. Those are all things that are generally considered to sound the same, assuming that they're all competently engineered, and operating within their limits, right?
I want to create a speaker cable switch that I can control from my chair. I can use relays, or I could run the speaker wires all the way to the chair and use a mechanical switch (yeah, relays are mechanical, but I'm talking something like a big blade switch that I would physically operate).
I can level match.
I can have my wife do the hookup of the cables so that I won't know what's what. She won't hang around, and she couldn't care less about which is which, as long as she hooks them up correctly.
I can do the math for the statistics, but I'll be happy to hear opinions from the people here on how many trials will be needed.
From there, what do I need to do to ensure that I have a proper test? I don't have any desire to cheat, or to have my more expensive gear come out the "winner". I also don't want to be accused of doing something wrong in the event that I can statistically differentiate the two.
Thanks.
Excellent. Since your wife is willing to help, I can make it fairly simple. Here are some guidlines you need to follow in order to get a valid test.
1. Level matching needs to be done with a voltmeter and levels should be matched to within .1 volts.
2. If you can build a switch box or comparator, that's great. If not, just shield your wife and your equipment from your vision and let her unplug and plug cables based on a random series. Remember, you need to level match every time whether you have a switchbox or not unless your switchbox is a comparator with its own volume control. I recommend you use a test tone from a CD for the level matching and then switch it to the music for the listening test. This needs to be done every iteration.
3. play exactly the same input material for exactly the same amount of time for each iteration.
4. Play it at normal listening level so that neither unit is approaching a clipping level.
5. Make sure your wife has the ramdomly developed series of iterations and you don't, assuming you are the listener.
6. Be sure she unplugs the cables and replugs them the same way each time. If the random series calls for there to be no change in the units, she still needs to go through the routine so that you aren't tipped off that the same unit playing this iteration.
7. Have her ask you whether you are listening to A or B. She should score right or wrong on the score sheet. Then you can tally the rights and wrongs for the statistics.
8. Do at least 10 iterations - 20 is better.
If you can get through all that, you will be able to say with authority that your results are valid. Enjoy the test and, please, post the results.
